Eric R.

WoZf

Kinship, Friendship, and Patron-Client Relations in Complex Societies

CORE AND PERIPHERY IN COMPLEX SOCIETIES

The anthropologists study of complex societies receives its major justification from the fact that such societies are not as well organized and tightly knit as their spokesmen would on occasion like to make people believe. If we analyze their economic systems, we shall find in any one such society resources which are strategic to the system - and organizations set up to utilize these strategic resources - but we shall also find resources and organizations which are at best supplementary or wholly peripheral. If we drew these relations on a map, some areas would show strong concentrations of strategic resources and the accompanying core organizations; other areas would appear in grey or white, economic terra incognita from the point of view of the larger system. The same point may be made with regard to political control. There are political resources which are essential to the operation of the system, and the system will try to remain in control of these. But there are also resources and organizations which it would be either too costly or too difficult to bring under direct control, and in these cases the system yields its sovereignty to competitive groups that are allowed to function in its entrails. I shall argue that we must not confuse the theory of state sovereignty with the facts of political life. Many organizations within the state generate and distribute and control power, in competition with each other and with the sovereign power of the state. As examples one might cite the Teamsters Union of the United States, the Mafia, or the American Medical Association. Thus we could also draw a map of political power for any complex society in which the key centers of control - Lenins strategic heights - appeared in red - showing strong concentrations of sovereign power 1

Eric R. Wou
while other political regions appeared as grey or white. We thus note that the formal framework of economic and political power exists alongside or intermingled with various other kinds of informal structure which areinterstitial, supplementary, parallel to it. Even the study of major institutions, such as of the American and German armies during World War II, or of factories in Britain and the United States, or of bureaucratic organizations, has yielded statements about the functional importance of informal groups. Sometimes such informal groupings cling to the formal structure like barnacles to a rusty ship. At other times, informal social relations are responsible for the metabolic processes required to keep the formal institution operating, as in the case of armies locked in combat. In still other cases, we discover that the formal table of organization is elegant indeed, but fails to work, unless informal mechanisms are found for its direct contravention, as in %he network of blat relationships among Soviet industrial managers. The anthropologist has a professional license to study such interstitial, supplementary, and parallel structures in complex society and to expose their relation to the major strategic, overarching institutions. In this paper, I should like to focus on three sets of such parallel structures in complex societies: kinship, friendship, and patron-client relations. Since my fieldwork experience has been confined to Latin-America and to the European Mediterranean, my examples will be largely drawn from these areas, and my thinking will be based largely on these examples. I shall indicate where I think it could be extended to other areas, but I shall expect to hear that they cannot be applied universally. We must not, of course, picture the structures of complex society as an ordered anarchy. The informal structures of which I have spoken are supplementary to the system: they operate and exist by virtue of its existence, which is logically, if not temporally, prior to them. Allow me to make use of Lewis Henry Morgans dichotomy of societm and ci~~itas to clarify my meaning. In societas, the principle of kinship embodies all or most strategic relations; in &&as, relations of political economy and ideology guide and curtail the functions of kinship. Let me caution that this is true more of kinship functions than of 2

Kinship,

Friendship,

and Patron-Client

Relations

kinship form. Indeed we are learning a great deal about just how far or how little kinship mechanisms can be stretched and bent to accommodate different interests. Nevertheless we must recognize a polarity in function. Relations may still have kinship form, but no longer primarily kinship functions. Take, for example, the corporate patrilineages in pre-Communist Southeastern China, studied by Freedman (1958). These units combined a kinship dogma of organization with the functions of commercial corporate organizations.
CORPORATE KIN CROUPS IN COMPLEX SOCIETIES

We may, reasonably, at the outset of our discussion of kinship in complex societies, ask when it is that we might expect to find kinship units of a corporate kind. There are two such units. One is the shallow local landed descent group, usually associated with primogeniture, of the kind which recently drew my attention in my study of the South Tyrolese (Wolf, 1962). Using a hypothesis put forth by Marshall Sahlins for the occurrence of similar groups in Polynesia (1957, pp. 294-295), I should argue that such units are likely to persist where the successful conduct of the enterprise requires the control - within one economic unit - of a number of ecological resouxces. In the case of the South Tyrolese, these resources would be agricultural land, meadowland close enough to the homestead to receive additional sources of fertilizer, pasture on higher ground, and forest. Division of the property upon inheritance would, in such circumstances, tend to splinter the viable economic unit into fragments, none of which could be meaningfully exploited by itself. The second kind of corporate kin unit for which we must account is the unilineal kinship corporation which transcends the local, three- or four-generation descent group. Thinking primarily of pre-Communist China and of the Near East, I would argue that such superlocal kinship corporations appear under two sets of conditions. The first of these concerns the mechanism regulating access to land. I would argue that where you gain access to land through paying rent, membership in a kinship coalition of the kind described would offer advantages 3

Eric R. Wolf
in increasing ones ability to obtain and keep land, and to affect the terms of rent. Second, and equally important, membership in a kinship coalition would be advantageous in situations where the state delegated the taxing power and the execution of other demands to entities on the local :level. Paying taxes through lineages or sub-lineages thus offers an opportunity to distribute the tax burden within the community on local terms, together with an ability to call on the protection and aid of these lineages. These two conditions, then, and perhaps also others which are not yet clear to me - the delegation of state fiscal power to entities lower down in the political hierarchy, coupled with the system which Hans Bobek (1962, pp. 233-240) has called rent-capitalism - would favor the emergence of the large-scale kin coalitions which anthropologists call ranked unilineal corporate descent groups.
CORPORATE COMMUNITIES

I would invoke similar factors for the continued existence, in certain parts of the world, of what I have labeled elsewhere the closed corporate peasant community (Wolf, 1955, 1957). Such communities - and I am thinking here primarily of Middle America, but also of Central Java, the Russian nzir, perhaps also the Near Eastern ~?ZU&XCG occur in areas where the central power does not or cannot intervene in direct administration, but where certain collective tasks in taxation and corvee are imposed on the village as a whole, and where the local village retains or builds administrative devices of its own natural and social resources. Both corporate kin groups and corporate peasant villages are growing fewer in the modern world. One is tempted to point out that historically the essential change in organizational forms leading from so-called traditional to modern societies lies in the elaboration - in the Mediterranean world - of non-agricultural corporate units like the muune and commenda which - though originally commercial or artisan kinship organizations developed the organizational potential of the corporate business struoture. Corporate kinship organization thus occurs where the groups 4

Kinship,

Friendship,

and Patron-Client

Relations

involved have a patrimony to defend, and where the interests associated with this defense can best be served by the maintenance of such a coalition. Such groups, too, must restrict and regulate the affinal bond, in order to restrict the number of people who may have access to the patrimony through inheritance. Another function served by such restrictions and regrilations of the affinal bond is to restrict the number of coalitions with other individuals that can be entered into by any one individual. The kinship coalition or the village coalition is thus made to override any coalitions which the individual may wish to form, by playing off affinal and consanguineal ties against each other.

INDIVXDUAL-~ENTERED

coALITIoNs

In situations where land and labor become free commodities, such corporate kin coalitions tend to lose their monopolies over resources and personnel. Instead, the individual is freed to enter into individual coalitions, to maximize his resources both in the economic field and in the marriage market. Increasing mobility, moreover, brings an increase in the number of possible combinations of resources, including varying combinations of knowledge and influence with access to goods or personnel. The theoretically unrestricted marriage market may thus be seen as offering increasingly wider choices of mates, thus providing the mechanisms for an increasing number of combinations of natural and social resources. In reality, however, the capacity to choose marriage mates is no more equal than the capacity to combine resources as commodities in the market. Theoretically, tycoon and beggarman may both have equal freedom to marry the kings daughter, just as both are free to sleep under the bridges of Paris. In actuality, however, we find that both access to resources and the capacity to maximize combinations through marriage relationships are unequally distributed throughout the social structure. Different potentials for effecting combinations of resources will result in a different functional load for the marital tie and for the mobilization of kin, and hence also in different patterns of marriage. In the Creole areas of Latin-America, as among the 5

Eric R. Wolf inhabitants of urban slums, we may find a minimal capacity to effect resource combinations reflected in a predominant or codominant pattern of matrifocal family arrangements. Among personnel located at the apex of society and capable of great potential in making resource combinations, we shall find corporate-like restrictions upon marital alliances to minimize the outward and downward flow of resources. In between, we shall encounter a whole range of patterns, representing more or less stable adjustments to possible combinations of goods, influence, knowledge, and power. Thus differential access to resources also leads to differences in the capacity for social maneuver, a differential capacity which is, in turn, refleoted in differential patterns of marriage choice. Seen from the perspective of resource distribution, the differential distribution of a population in terms of resources has been called the class system of a society. Seen from the perspective of the anthropologist interested in kinship, overlapping circles of kin tend to cluster in what one might call kinship regions. To the extent that kinship bonds constitute one set of resources for an individual or a family, the distribution of kinship alliances forms one important criterion for demarcating the classes of a society. As Schumpeter has said, the family, not the physical person, is the true unit, of class and class theory (1955, p. 113). In this regard, anthropologists need to pay much more attention to the rise and fall of families than they have done in the past. The best material to date comes from China, where a number of studies show the rise of families to gentry status, as well as their subsequent decline (see, for instance, Fei, 1953; Hsu, 1948; Yang, 1945). Similarly, Pi-Sunyer has recently shown how in the Mexican town of Zamora a new elite of entrepreneurs, who rose by their boo&traps during the revolution to displace an older landed aristocracy, has nevertheless fathered a set of sons who - in the changed oircumstances of their lives model themselves on that older aristocracy, to the detriment of the parental enterprises created by their self-educated and unpolished fathers (1962). I have, similarly, described how in Puerto Rico poor immigrants from Spain rose from rags to riches in the course of an exploitative process, but how the sons
6

Kinship,

Friendship,

and Patron-Client

Relations

of these immigrants did not take up the parental enterprise. Instead, the father would send home to Spain for a poor young kinsman or youth from his home community, discipline him mercilessly in the tasks of business, turn him into a son-in-law, and pass the business on to him, rather than to the no-good sons may follow (Wolf, 1956). Here, too, the anthropologist Schumpeters lead and ask himself why and how some families rise and others fall, quite apart from accidents, as he says, to which we attribute a certain importance but not the crucial role (1955, p. 118).
PERSISTENT FUNCTIONS OF THE FAMILY

Nor is it at all self-evident, to this writer, why families - rather than some other kind of unit - should be the functional entities within kin circles and in connecting circles. If we do not regard the family as a natural group, then we must at least assay its functional capacity and range, to account for its continued persistence. One of its characteristics which continues to recommend it is its ability effectively to unite a number of functions. There are, of course, the usual functions of economic provisioning, sooialization, the exchange of sexual services, the bestowal of affect. Although each of these functions could be handled in segmented and institutionalized fashion by a separate institution, the family can perform these multiple tasks in small units of output and in quick succession, with a relatively low cost and overhead. At any one time, the demands of a family represent only small-scale demands, for a quart of milk rather than for a railroad car, a song rather than a jukebox, an aspirin rather than the output of Lever Brothers. Moreover, these small-scale demands occur in quick succession, and involve a rapid shift of labor to meet them, a trip to the store to get a bottle for the baby when the old one breaks, followed by the preparation of peanut butter and jelly sandwiches, followed by a game of chess. Maximally efficient for the least amount of cost, therefore, the family is also maximally adaptive to changes in the conditions that define and circumscribe its existence. This is especially important, I believe, in families with meager resources
D 7

Eric R. Wolf
where labor can be increased to meet variable demands - as when a man takes over an extra job to pa.y for a refrigerator or when the wife tends a sick baby all night - without incurring expenses other than the exploitation of self. Here we may also underline the fact that in its pursuit of multiple purposes, the family remains the multi-purpose organization par exce.ZZence in societies increasingly segmented into institutions with unitary purposes. As such it may have compensatory functions, in restoring to persons a wider sense of identity beyond that defined by unitary demands of the job, be this cutting cane on a Puerto Rican plantation or tightening nuts on bolts on an assembly line. Let me make an additional point, however. It is notable that a relation continues to exist between the way in which a family carries out these multi-purpose tasks and the ways in which it is evaluated in the eyes of the larger community. The family not only performs all the tasks we have just described; it remains also, even where ties of kinship are highly diffuse, the bearer of virtue, and of its public reflection, reputation. Because the family involves the whole man, public evaluations of a man are ultimately led back to considerations of his family. Moreover, any gross infringement of virtue by one of its members reflects on the amount of virtue held by the others. This virtue has two aspects, one horizontal, in relationship to class equals, one vertical or hierarchical, in relation to class groups above and below ones station. The horizontal aspect of virtue refers to the guarding of a familys reputation in equivalent relation to other peoples reputations. Standards for evaluating reputations are culturally highly variable; yet in each society there exist vital indices for the relative ranking of reputations. These rankings define whom one can trust, whom one may marry. Invariably, they refer back to ways in which people handle their domestic affairs. Frequently, as in the EuropeanMediterraneanoramong theLadinos of Latin-America, reputation is tied to what is potentially its weakest link, the sexual behavior of ones womenfolk. The concept of honor, in its horizontal aspect, implies a fixed amount of reputation for each contestant in the game of honor, an amount which can be lessened or increased in competitive interaction with others.

Kinship,

Friendship,

and Patron-Client

Relations

Such interaction establishes ones social credit rating, a rating in which intrafamilial behavior is the final referent. Moreover, past familial behavior has important bearing on present and future evaluation. This element is sometimes missed in the discussion of societies characterized by bilateral kinship arrangements. The maintenance of a family name, the importance of family names, even in situations where genealogical reckoning is weak or shallow, make less sense when thought of in terms of patrilineal or matrilineal filiation than in terms of the storage and enlargement of virtue for each family. What has been said here of horizontal virtue holds with increased intensity for members of ranked class groups. The point is obvious and need not be labored in this context.
COOPERATION OF KIN IN NON-KIN SITUATIONS

Not only does filiation with a family define ones social credit rating. It also structures the nature of social resources at ones command in operations in the non-kin realm. Kin relations in such maneuvers possess two advantages over non-kin ties. First, they are the product of social synchronization achieved in the course of socialization. The private relation of trust may thus be translated into cooperation in the public realm. I would like to point here, for example, to the relations of uncles and nephews in Euramerican culture which gave rise to the concept of nepotism. It is interesting, parenthetically, that this relation is described in great detail in such sources as the French chmzsons de geste, including all the psychological attributes ascribed by Homans and Schneider to the relation of mothers brother and sisters son, in contrast to the relation between son and father, in the absence of known patterns of complementary filiation and matrilateral cross-cousin marriage. Moreover, such a relation between kinsmen can rely on the sanctions of the kin network, as well as on the sanctions of the public realm. Should one partner to the relationship fail in his performance, his alter can mobilize against him not only the immediate sanctions of the eg,o-alter tie, but all the other bonds that link ego and alter to other kin. It is obvious, of course, that such a reliance on kin may also entail liabilities to one or the other member of the 9

Eric R. Wolf partnership. Kinsmen may become parasitic upon one another, thus limiting the capacity of any one member to advance his
wealth or power. The clearest gain from such a relation should therefore appear in situations where publiclaw cannot guarantee adequate protection against breaches of non-kin contracts. This can occur where public law is weak, or whereno cultural patterns of cooperation between non-kin exist to guide therequired relationship. It can alsooccur in dealings which border on the illegal or extra-processual. Cooperation among kin, for example, is important in gangster organizations, even where non-kin relations may sometimes be forced at gun point, or in political hatchet-work, in which kin relations are emljloyed privately to prune the political underbrush. It is finally useful for kin to cooperate where access to the law would entail such costs and complications as to leave the partners to a dispute economically or otherwise deprived after settlement. The relation of kin in non-kin operations, therefore, implies a clear balance of gains and costs, in which the gains outweigh the costs only when cooperation with non-kin is clearly more hazardous and disadvantageous. KINDS OF FRIENDSHIP

-, -1 -?. _.:

At this point, the tie of kinship merges with the tie of friendship. In contrast to the kin tie, the primary bond in the friendship dyad is not forged in an ascribed situation; friendship is achieved. If we are to make headway in a sociological analysis of the friendship tie, me must, I believe, distinguish two kinds of friendship. I shall call the first expressive or emotional friends&p, the -.__ __- second +s~rumental. friendship:. From the point of -view of the friendship dyad, emotional friendship involves a relation between an ego and an alter in which each satisfies some emotional need in his opposite number. This is the obviously psychological aspect of the relation. Yet the very fact that the relation satisfies a deficit of some kind in each participant should alert us also to the social characteristics of the relation involved. It leads us to ask the question: under what kind of conditions can one expect to find an emotional deficit in two persons which draws them into the relation described?
-; 10

I : :.

Kinship,

Friendship,

and Patron-Client

Relations

Here it is useful to look upon friendship as a count.ervailing force. We should, I think, expect to find emotional friendships primarily in social situations where the individual is strongly embedded in solidary groupings like communities and lineages, and where the set of social structure inhibits social and geographical mobility. In such situations, egos access to resources natural and social - is largely provided by the solidary units; and friendship can at best provide emotional release and catharsis from the strains and pressures of role-playing.
FRIENDSHIP: A MIDDLE-AMERICAN CASE

I think primarily here, for instance, in terms of my own experience, of the behavior of Indians in closed corporate communities in Middle America. The community is solidary towards outsiders and against the outside; it maintains a monopoly of resources - usually land - and defends the first rights of insiders against outside competition. Internally, it tends to level differences, evening out both the chances and the risks of life. This does not lead to the warm communal relations sometimes imputed to such a structure. Quite the contrary, we may note that envy and suspicion play an essential part in maintaining the rough equality of life chances. Friendship in such a community provides an escape from the press of life, but it does not in and of itself serve to alter the distribution of resources. Ruben Reina has described how friendship works in such a community in Guatemala. For the Indians, he says, it offers an emotional fulfillment and a means of assuring oneself that one will not be standing alone. Before marriage and after childhood, the camarada complex reaches high emotional intensity at that transition in life when a Chinautleco achieves adult status but has not acquired all its emotional rewards. At the same time, the very intensity of the relation has a tendency to dissolve it. The explanation seems to lie in the fact that Indians seek extreme confidence (con$anza) and this in itself endangers friendship. They demand reciprocal affection, and it is expected that the camarada will act only in a manner which will bring pleasure to his friend. Camarudas are jealous of each ot,her:

11

Eric R. Wolf
once a high intensity of friendship was attained, scenes of jealousy and frustration could be expected and the cycle would end in a state of enmity. Hence such emotional friendship is also ambivalent. As Reina says, they are proud of this relationship and affectionate in it, but from a practical viewpoint have mixed feelings. A camarada is a potential enemy when the puesto [prescribed role and status] is lost. A certain reserve on the part of the cama/radas is therefore observed, especially in the realm of family secrets, plans, and amounts earned at work. Friendship is maintained not for economic, political, or practical purposes, but only an emotional fulf?llment (Reina, 1959). Emotional friendship is thus self-limiting; :its continuation is threatened from the inside. It is also subject to limitation from the outside. Here we may use Yehudi Cohens observation that solidary groups feel cross-cutting friendship ties as a threat and hence will attempt to limit them. He advances this hypothesis to explain the institution of the inalienable friend in what he calls maximally solidary communities, characterized in the main by corporate kin groups (1961, p. 375). In contrast to emotional friendship is what I have called instrumental friendship. Instrumental friendship may not have been entered into for the purpose of attaining access to resources - natural and social - but the striving for such access becomes vital in it. In contrast to emotional friendship, which : restricts the relation to the dyad involved, in instrumental friendship each member of the dyad acts as a potential connecting link to other persons outside the dyad. Each participant is a sponsor for the other. In contrast to emotional friendship, which is associated with closure of the social circle, instrumental friendship reaches beyond the boundaries of existing sets, and seeks to establish beachheads in new sets. Ruben Reina, whose Indian material I have described, contrasts the Indians in Chinautla with the Ladinos. To the Ladinos, friendship has practical utility in the realm of economic and political influence; this friendship is looked upon as a mechanism beneficial from the personal viewpoint.
12

Kinship,

Friendship,

and Patron-Client

Relations

Cwello, a favorite expression among the Ladinos, indicates that a legal matter may be accelerated, or a job for which one is not totally qualified might be secured through the personal influence of an acquaintance who is in power or knows a third party who can be influenced. The cu,eZZo complex depends upon the strength of friendship established and is often measured in terms of the number of favors dispensed to each other. It finds its main support in the nature of a convenient social relationship defined as friendship. It follows that, for the Ladinos of Chinautla, the possession of a range of friends is most favorable (Reina, 1959, pp. 44-45). Despite the instrumental character of such relations, however, a minimal element of affect remains an important ingredient in the relation. If it is not present, it must be feigned. When the instrumental purposes of the relation clearly take the upper hand, the bond is in danger of disruption. One may speculate about the function of this emotional burden. The initial situation of friendship is one of reciprocity, not of the tit-for-tat kind which Marshall Sahlins has referred to as balanced reciprocity, but of more generalized reciprocity. The relation aims at a large and unspecified series of performances of mutual assistance. The charge of affect may thus be seen as a device for keeping the relationship a relation of open trust or open credit. Moreover, what may start out as a symmetrical reciprocal relationship between equal parties may, in the course of reciprocal services, develop into a relation in which one of the parties - through luck or skillful management - develops a position of strength, the other a position of weakness. The charge of affect which retains the character of balanced reciprocity between equals may be seen as a device to ensure the continuity of the relationship in the face of possible ensuing imbalance. Hence, too, the relation is threatened when one party is too clearly exploitative of the other (Pitt-Rivers, 1954, p. 139). Similarly, if a favor is not forthcoming, the relation is broken and the way is left open for a realignment of friendship bonds. The relation thus contains an element which provides sanctions internal to the relation itself. An imbalance in the relation automatically severs it. 13

Eric R. Wo.$f
CORPORATE GROUPS AND MIGRANT POPULATIONS

Just as the persistence of corporate groups in a society discourages the mobilization of friendship ties for mobility beyond the corporate group, so it also places a special restriction on the use of kin bonds to effect this crossing of social boundaries. I believe this to be characteristic of the closed corporate communities of Middle America. There the individual who wishes to move beyond the orbit of the community - or is pushed beyond that orbit - is frequently accused of actual or potential witchcraft, and thus defined as a deviant, against whom social sanctions may be invoked. This can be seen most clearly, of course, in witchcraft accusations. Manning Nash has given us a remarkable and convincing picture of how witches in the corporate community of Amatenango are socially isolated, until their kinfolk abandon them to their ultimate fate of death (Nash, 1960). The records of the Cbiapas project of the University of Chicago are full of cases of splinter groups which have left the villages of their origin under the onus of witchcraft accusations to settle elsewhere. When a person migrates from such a community, he is lost to it unless the corporate mechanisms break down and allow him to resume relations with kinfolk in the village, or further migrants seek his help in the greater outside. Similarly, in the South Tyrolese village I studied, the prevalent pattern of inheritance by one son breaks up the sibling group and causes the supernumary siblings, die weichenden Erben, the yielding heirs, to emigrate. In such cases, contact between remaining heir and migrants is cut and lost.

OPEN

ORGANIZATION

AND

MIGRANT

POPULATIONS

This is not, however, the case in open communities where neither corporate communal organization nor/and corporate lineal groups divide potential stay-at-homes from potential migrants. There a person is free to mobilize both friendship and kinship ties to advance his mobility both inside and outside the community. Kinship ties with migrants are not lost - they become valuabIe assets for the transmission or distribution of goods and services. Thus, the Buerto Ricans of San Jose retain
14

Kinship,

Friedship,

and Patron-Client

Relations

strong ties with their migrant kin in San Juan and in the United States. The people of Tret, the Italian community which
I studied contrastively with St Felix, the German South

Tyrolese community, keep track of every relative who has gone to the United States, and keep in touch through letters and
mutual gifts. And Ernestine Fried1 has shown in her study of

Vasilika in Boetia that the role of kinship ties as a mechanism

for maintaining urban-rural connections is extensive and permeating. Nor does a change in social status from poorer to wealthier Greek peasant, or to any other more prestige-giving position, result in a rupture of kinship ties and obligations (1959, p. 31). Finally, it will have been noted that the instrumental friendships discussed above thrive best in social situations which are relatively open, and where friends may act as sponsors for each other in attempts to widen their spheres of social maneuver. The twentieth century has, however, also witnessed a new form of social closure, not this time on the level of the landed corporate group, but in the tendency of large-scale bureaucratic organizations to lessen the area of free maneuverability. In such large bureaucracies as industrial concerns or armies, instrumental friendship merges into the formation of cliques or similar informal groups.
CLIQUES

Compared to the type situation discussed above, in which the friendship relation still covers the entire role repertoire of the two participants, clique friendship tends to involve primarily the set of roles associated with the particular job. Nevertheless, the clique still serves more purposes than are provided for in the formal table of organization of the institution. It is usually the carrier of an affective element, which may be used to counterbalance the formal demands of the organization, to render life within it more acceptable and more meaningful. Importantly, it may reduce the feeling of the individual that he is dominated by forces beyond himself, and serve to confirm the existence of his ego in the interplay of small-group chit-chat. But it also has important instrumental functions, in rendering an unpredict15

Eric R. Wolf

able situation more predictable, and in providing for mutual support against surprise upsets from within or without. This is especially true in situations characterized by a differential distribution of power. Power superiors and inferiors may enter into informal alliances to ensure the smooth prosecution of their relationship, to guard against unbidden inquiries from the outside or competition from the inside, to seek su.pport for advancement and other demands. Prize examples of such informal alliances are provided by J. Berliners discussion of familyness and bZat, influence, among Soviet industrial managers (1957); but they can be had in any account of the functioning of a large bureaucratic organization. Indeed, paraphrasing a comment of Edward Shils, an interesting perspective on the study of such large organizations may be gained by looking upon them as organizations of supply for the cliques they contain, rather than the other way round, by visaalizing the clique group as a servant of the bureaucracy that provides its matrix.
PATRON-CLIENT RELATIONS

When instrumental friendship reaches a maximum point of imbalance so that one partner is clearly superior to the other in his capacity to grant goods and services, we approach the critical point where friendships give way to the patron-client tie. The relation between patron and client has been aptly described as lop-sided friendship (Pitt-Rivers, 1954., p. 140). As in instrumental friendship, a minimal charge of affect invests the relation of patron and client, to form that trust which underwrites the promise of future mutual support. Like kinship and friendship, the patron-client tie involves multiple facets of the actors involved, not merely the segmental needs of the moment. At the back of the material advantages to be gained by the client, says Kenny of patron-client relations in Spain, there lies not only a striving to level out inequalities but also a fight against anonymity (especially in the urban setting) and a seeking out of primary personal relationships (1962, p. 136). The two partners to the patron-client contract, however, no longer exchange equivalent goods and services. The offerings of the patron are more immediately tangible. He provides economic 16

Kinship,

Friendship,

and Patron-Client

Relations

aid and protection against both the legal and illegal exactions of authority. The client, in turn, pays back in more intangible assets. These are, first, demonstrations of esteem. The client has a strong sense of loyalty to his patron and voices this abroad. By doing so, he constantly stimulates the channels of loyalty, creates good will, adds to the name and fame of his patron and ensures him a species of immortality (Kenny, 1962, p. 136). A second contribution by the client to his patron is offered in the form of information on the machinations of others. A third form of offering consists in the promise of political support. Here the element of power emerges which is otherwise masked by reciprocities. For the client not only promises his vote or strong arm in the political process, he also promises - in effect - to entertain no other patron than the one from whom he has received goods and credit. The client is duty-bound not merely to offer expressions of loyalty, but also to demonstrate that loyalty. He becomes a member of a faction which serves the competitive purposes of a faction leader. Crises, says Kenny, clearly reveal this when protestations of loyalty and support significantly show the alignment of different patronage forces. It is this potential competition of patron with patron that offers the client his leverage, his ability to win support and to insist on its continuation. The relation remains reciprocal, each party investing in the other.
VARIATIONS IN PATRON-CLIENT TIES

We may, moreover, engage in some speculation as to the form which the patron-client relation will take in different circumstances. I should expect the relation here analyzed to occur where no corporate lineal group or corporate village intervenes between potential client and potential patron, but where the network of kin and friendship relations is sufficiently open for each seeker after support and each person capable of extending support to enter into independent, dyadic contracts (Foster, 1961). Moreover, such ties would prove especially functional in situations where the formal institutional structure of society is weak and unable to deliver a sufficiently steady supply of goods and services, especially to the terminal levels of the social order. 17

Eric R. Wolf Under such conditions, there would be customers for the social insurance offered by potential clients, while the formation of a body of clients would increase the ability of .patrons to influence institutional operation. These considerations would lead one to predict further that patron-client relations would operate in markedly different ways in situations structured by corporate groups, or in situations in w&h the institutional framework is strong and ramifying. Among the South Tyrolese, there is no patron-client tie of the kind discussed here. Its place is taken by political party leadership which communicates hierarchically to the various lineal corporate units in the village. On the other hand, where you have super-local unilineal descent groups, as in China and the Near East, we find the patron incorporated into the lineage, in the person or persons manning the executive gentry positions in the lineage. Similarly, among corporately organized Indians in Middle America, the individual can approach a patron - hacienda owner or political power figure only as a member of the group, and the patron then acts as power broker relating the entire group to the institutional framework outside it. On the other hand, where there are no corporate kin or village units of the type indicated, but where the institutional framework of society is far flung and solidly entrenched, patronage cannot lead to the formation of bodies of followers relatively independent of the formal structure. Rather, patronage will take the form of sponsorship, in which the patron provides connections (hence the Spanish enchfe - plug-in) with the institutional order. In such circumstances, his stock-in-trade consists less of the relatively independent allocation of goods and services than of the use of influence. Correspondingly, however, his hold on the client is weakened, and in place of solid patron-client blocks we may expect to encounter diffuse and cross-cutting ties between multiple sponsors and multiple clients, with clients often moving from one orbit of influence to another.
THE PROBLEM OF NATIONAL CHARACTER

I cannot refrain, at the end of this discussion, from pointing out a point of encounter with what has sometimes been called the 18

Kinship,

Friendship,

and Patron-Client

Relations

national-character approach. When one examines the work of Benedict, Mead, and others who have devoted their attention to the problem of defining national character, one is struck by the fact that they have utilized - in the main - data on the interpersonal sets discussed in this paper, and on the etiquettes and social idioms governing them. Take, for instance - and picking at random - Geoffrey Gorers account of the intricacies of mate selection involved in the American dating complex (1948), or Benedicts discussion of the circle of on and giri obligations between persons of different hierarchical status (1946), or Rhoda Metrauxs analysis of the constitution of the French foyer (1954). There is no need to labor this point. It is obvious that such descriptions and analyses do not cope with the institutional features of national structure. Yet it is equally possible that complex societies in the modern world differ less in the formal organization of their economic or legal or political systems than in the character of their supplementary interpersonal sets. Using the strategy of social anthropology, moreover, we would say that information about these sets is less meaningful when organized in terms of a construct of homogeneous national character than when referred to the particular body of social relations and its function, partial or general, within the supplementary or parallel structure underlying the formal institutional framework. If our argument is correct that these supplementary sets make possible the functioning of the great institutions, then it must also be true that these supplementary sets developed or changed character historically, as the great institutions developed historically. And with changes in these supplementary sets we should also expect to find changes in the norms governing these sets, and in the symbolic forms assumed by these norms. The integration of the great society requires the knitting of these interstitial relations. As the integration of society is promoted by certain groups who draw after them a variety of others, some groups moreover set the pace and tone in the formation of the new patterns, which draw in or influence the segmental patterns of other groups. The patterns of interpersonal etiquettes of one group are then recut and reshaped to fit the patterns of interpersonal etiquettes utilized by the 19

Eric R. Wolf
tone-setting group. Put in terms of reference theory, we might say that the choice of behavioral etiquettes and the direction of their circulation reflect the degree of dominance of one or another reference group within the society. An example of downward circulation of such patterns would be the spread of courtly forms in France (Elias, 1939), the establishment and diffusion of public-school manners in Britain, the communication of urban forms to rural groups via the kinship network in Greece and Italy (Friedl, 1959; Wolf, 1962). But there can also be cases of upward circulation of behavioral models, reflecting changes in the distribution of power in a society, as when the etiquette governing the relation of traditional hacienda owner and agricultural worker in Puerto Rico was transferred to pattern the relation between the new island-wide political leadership and its mass following (Wolf, 1956, pp. 212-213), or when the behavioral etiquette of a despised interstitial group in Mexico became the behavioral grammar standardizing interaction between power-seekers and followers (Wolf, 1959, ch. 11). Description and analysis of the supplementary interpersonal sets discussed in this paper thus not only reveal a great deal about the hidden mechanisms of complex society. Description and analysis of the origin and circulation of the models of etiquette structuring these sets also reveal much of the social dynamic, of the changing distribution of forces in the social body. If such studies do not lead us to definitions of national character, as this term has hitherto been employed, they nevertheless indicate the way in which the parallelogram .of social forces in one society differs from that of another.
ACKNOWLEDGEMENT Thanks are due to the Editor of the American Anthropologist for permission to quote a passage from Two Patterns of Friendship in a Guatemalan Community by Ruben Reina.

REFERENCES
RUTH F. 1946. The Chrysanthemum and the Sword. Boston: Houghton Mifflin. BERLINER, JOSEPH. 1957. Factory and Manager in the U.S.S.R. Cambridge, Mass.: Harvard University Press. BENEDICT,

20

E&ship,
B OBEK,

.Z%iendship, and Patron-Client

Relations

HAN s. 1962. The Main Stages in Socioeconomic Evolution from a Geographic Point of View. In Philip L. Wagner & Marvin W. Mikesell (eds.). Readings in Cultuml Geography. Chicago: University of Chicago Press, pp. 218-247. COHEN, YEHUDI. 1961. Patterns of Friendship. In Yehudi Cohen (ed.), Social Structure and Personality: A Casebook. New York: Holt, Rinehart & Winston, pp. 351-386. ELIAS, NORBERT. 1939. Ueber den Prozess der Zivilisation: Soxiogenetische und Psychogenetische Untersuchungen, 2 ~01s. Basel: Verlag Haus zum Falken. FEI, HSIAO-TUNQ. 1953. Chinas Gentry. Chicago: University of Chicago Press. FOSTER GEORGE FM. 1961. The Dyadic Contract: A Model for the Social StructuGe of a Mexican Peasant Village. American AnthropoEogist 63: 1173-1192. FREEDMAN, MAURICE. 1958. Lineage Organization in Southeastern China. London School of Economics Monographs on Social Anthropology No. 18. London: Athlone Press. FRIEDL, ERNESTINE. 1959. The Role of Kinship in the Transmission of National Culture to Rural Villages in Mainland Greece. American Anthropologist 61: 30-38. CORER, GEOFFREY. 1948. The American People. NewYork: Norton. HSU, FRANCIS L. K. 1948. Under the Ancestors Shadow: Chinese Culture and Personality. New York: Columbia University Press. KENNY, MICHAEL. 1962. A Spanish Tapestry: Town and Country in Cast&. Bloomington: University of Indiana Press. M~TRAUX, RHODA. 1954. Themes in French Culture. In Rhoda Metraux & Margaret Mead (eds.), Themes in French Culture: A Preface to a Study of French Community. Hoover Institute Series, D: Communities, No. 1. Stanford: Stanford University Press, pp. l-65. NASH, MANNING, 1960. Witchcraft as Social Process in a Tzeltal Community. Am&ica Indigena 20: 121-126. PI-SUNYER, ORIOL. 1962. Personal communication. PITT-RIVERS, JULIAN A. 1954. The People of the Sierra. New York: Criterion Books. REINA, RUBEN. 1959. Two Patterns of Friendship in a Guatemalan Community. American Anthropologist 61: 44-50. SAHLINS, MARSHALL D. 1967. Differentiation by Adaptation in Polynesian Societies. Journal of the Polynesian Society 66: 291-300.

21

Eric R. Wolf
JOSEPH. 1955. Social Glasses. Imperia&m: Two Essays. New York: Meridian Books. WOLF, ERIC R. 1955. Types of Latin American Peasantry. American Anthropologist 57: 452-47 1. 1956. San Jose: Subcultures of a Traditional Coffee Municipality. In Julian Steward (ed.). The People of Puerto Rico. Urbana: University of Illinois Press, pp. 171-264. 1957. Closed Corporate Peasant Communities in Mesoamerica and Central Java. Southwestern Journal of Anthropology 13: 1-18. 1959. Sons of the Shaking Earth. Chicago: University of Chicago Press. 1962. Cultural Dissonance in the Italian Alps. Comparative Studies in Society and History 5: 1-14. YANG, MARTIN c. 1945. A Chinese Village: Taitou, Shantung Province. New York. Columbia University Press. SCHUMPETER,

Addendum After completion of this paper, I encountered in an excellent discussion, based on African materials, many ideas on the subject of friendship which parallel my own. This is
GIBBS, JAMES I,., Jr. 1962. Compensatory Blood-Brotherhood: A Comparative Analysis of Institutionalized Friendship in Two African Societies. Proceedings of the Minnesota Academy of Science 30: 67-74.

22